Download Gene and Body - Crowley Davis Research, Inc.

Gene and Body: Building and Maintaining the Phenotype of Living Organisms
Tim Otter and Robert Davis
Crowley Davis Research
280 South Academy Ave., Suite 140
Eagle, ID 83616
[email protected]
Abstract
Biological development is a progressive process, building
from the fertilized egg via the embryo to the adult body.
Development involves coordinated, stage-specific expression of genes, but the unidirectional, 1:1 mapping of
genotype onto phenotype implied by the “central dogma”
(DNA Æ RNA Æ protein) is not adequate to explain what
happens during development. This paper explores the global
relationship between genotype and phenotype in a manner
that places the central dogma in the context of living cells.
Introduction
In simple terms, development is the process of becoming
an adult. Modern developmental biology –the study of
those phenomena, processes, and mechanisms that lead
from fertilized egg via embryo to adult– involves the
fusion of two disciplines, embryology and molecular
genetics. One of the main goals of developmental biology
is to understand how genes shape the adult body.
Recently, as technologies for large-scale sequencing of
DNA have come available, the composition and
organization of entire genomes (human, mouse, pufferfish
(Fugu), nematode worm (Caenorhabditis elegans), fruit
fly, plant (Arabidopsis thaliana), and others) have been
determined. This broadening of perspective has led, in
turn, to pressing needs for computational approaches to
handle large (genomic) data sets and for a theoretical
framework for biology that can help to model the process
of development in quantitative terms.
Merging of biology with computer science is a two-way
street: biologists stand to gain more powerful and
quantitative modeling tools while computer scientists
benefit by incorporating biological principles to produce
more robust and fault-tolerant computational strategies.
However, until recently biologists have not generally
framed their questions in this way, and so bringing these
two disciplines together has revealed some problems and
inadequacies in the conceptual framework of biology. In
the interest of promoting fuller collaboration, this paper
explores the global relationship between genotype and
phenotype, between the information contained in an
organism’s genome and the processes for building and
maintaining a multicellular body.
As explained below, the key is understanding the cellular
context in which the genes operate. Establishing and
maintaining order requires energy, and part of what a cell
provides is a supply of energy to support its high degree of
organization. The phenotype is a complex, spatially
ordered and temporally defined state, but the genotype is
essentially a parts list. In the words of Franklin Harold
(2001), it is best to think of a cell “…as a spatially
structured self-organizing system made of gene-specified
elements. Briefly, the genes specify What; the cell as a
whole directs Where and When…”.
One Gene, One Protein?
Beadle and Tatum’s classic (1941) studies on metabolic
pathways established the fundamental 1:1 correspondence
between the genetic unit, a gene, and a phenotypic
property, a functional enzyme. Subsequently their “one
gene, one enzyme” hypothesis has been modified and
updated to allow for multi-subunit enzymes, proteins
without enzymatic activity, and splicing together of coding
regions by removal of introns, but Beadle and Tatum’s
essential concept –the correspondence between gene and
enzyme activity– remains.
Soon after Watson and Crick proposed the double helix
model of DNA (1953) the genetic code was deciphered,
and by 1967 the collinear nature of genes and proteins
became apparent. Thus, the correspondence between
genes and proteins holds at the level of their building
blocks, whereby 1 codon, a triplet of DNA nucleotides,
specifies 1 amino acid in the corresponding protein. This
relationship is summarized in the central dogma of
molecular biology:
DNA Æ RNA Æ protein
or more simply,
gene Æ protein
As long as the encoding gene has the proper sequence of
nucleotides, the resulting polypeptide will have the correct
sequence of amino acids. The ultimate, functional shape of
the properly folded polypeptide chain depends, in turn, on
its sequence of amino acids. Folding involves local
interactions between neighboring amino acids to produce
α-helixes and β-sheets, which associate to form higher
order domains. So, in a sense, for at least some proteins,
the genetic information (genotype) specifies, through the
folding process, a protein’s shape and thereby what kinds
of complementary shape(s) that protein can bind to (its
function or phenotype).
Mutations that change the
sequence of amino acids so as to alter the folded protein’s
shape thereby impair (very rarely, enhance) its function.
In sum, at the level of individual genes that code for
single-subunit enzymes, there is a 1:1 correspondence
between a gene and the function carried out by the encoded
protein. By extrapolation, then, one might conclude that
the phenotype, which includes the organism’s physical
traits, metabolic state, stage of development, and other
discernable characters, is simply the aggregate of all
expressed genes in a cell, and that every trait in the
organism is therefore specified in the genome (in its DNA).
However, things are not so simple. As the following
examples illustrate, function often derives from complex
interaction of proteins encoded by different genes.
For many proteins, the correct amino acid sequence
alone is not sufficient to produce the functional, folded
shape. There are several reasons for this. In some cases,
the translated amino acid sequence actually specifies an
inactive precursor (e.g., proinsulin), and activation may
involve cutting by an enzyme or covalent modification of
one or more amino acids (e.g., phosphorlyation or
acetylation). As a result, we could not hope to infer the
function of such a protein simply on the basis of its
(genetic) sequence; functionality depends on a protein
specified by another gene.
To fold properly, some proteins require other proteins
called molecular chaperones to help them along. Folding
of tubulin, for example, is guided by a complex of
chaperones called TRiC or CCT (Leroux and Hartl 2000).
Tubulin is found only in eukaryotic cells, where its role is
to assemble into rod-shaped polymers called microtubules
that help to establish cell polarity, maintain cell shape, and
move chromosomes during mitosis.
Assembly of
microtubules requires precise, complementary molecular
fit of the tubulin subunits. Ill-formed tubulin does not
assemble. That is why, when tubulin genes are cloned in
bacteria (prokaryotes), tubulin protein is made but no
microtubules form: bacteria lack the chaperones needed to
render tubulin assembly-competent.
Thus, the same gene, one that functions properly in its
natural eukaryotic setting, becomes functionally bankrupt
in a prokaryotic cell. Clearly, tubulin’s function is not an
exclusive property of the encoding gene, but it is defined in
part by the cellular context.
Most proteins do not function in isolation but instead are
components of macromolecular machines (e.g., ribosomes,
membranes, or complexes of metabolic enzymes) whose
function integrates the activity of several different proteins
or segments of RNA that are encoded by different genes.
The replication fork machine is a macromolecular complex
that performs one of the most fundamental tasks in living
cells: it copies the DNA, so that during cell division each
daughter cell receives a full set of genetic information.
DNA replication actually involves many steps and
processes, all of which must be coordinated so that the
machine copies both strands of the DNA.
DNA
polymerase, the enzyme that copies and corrects its own
errors (“proofreads”) moves in one direction only along the
template strand of DNA. Since the two strands of the
double helix run antiparallel, to copy the other strand the
replication fork machine makes a series of discontinuous
fragments and then links them with another enzyme, DNA
ligase. When all is said and done, more than a dozen
proteins, several of these with multiple subunits, comprise
the replication fork machine (Table 1).
Protein
Function
Subunit Structure
ORC
Helicase
Gyrase
SSB
Primase
Clamp
Loader
DNApol
Ligase
RNAase
Recognize start (origin)
Unwinds DNA helix
Relieves supercoil tension
Stabilizes ssDNA
Synthesizes RNA primer
Slides DNApol along helix
Loads & positions clamp
Copies & proofreads
Seals fragments
Removes primer
multiple monomers
6-mer
4-mer: 2 x 2 subunits
multiple monomers
4 subunits
ring-shaped dimer
5 subunits
3 subunits
1 subunit
1 subunit
~5 more
Repair & mitochondrial replication
Table 1. Molecular composition of a generic machine for
replication of DNA. The names of proteins vary from species
to species, but their functions (middle column) are conserved.
Thus, to claim that all cells need to copy their DNA is
DNA polymerase would be absurd. The cell must first
unwind the helix (helicase), relax the resulting supercoil
twists (gyrase), stabilize the single stranded DNA (SSBs),
copy and proofread (DNA polymerase), start the fragments
along the opposite strand with an RNA primer (primase),
link the fragments (ligase), and so forth.
The bottom line is that for some proteins, one of the
most fundamental aspects of phenotype, activity or
function, cannot be inferred from the information in a
single gene. We must also know about the modifications
to the protein’s structure and function that occur in the
living cell, and we must understand the context in which
the protein functions.
In summary, a protein’s function, its phenotype, is the
product of the evolution of that protein’s gene sequence
plus the evolution and function of any other genes that are
required to render that protein fully active. Therefore, to
generalize from the simplest case where no such controls
happen to apply (the protein folds spontaneously into its
active conformation, is not modified after translation, and
is not a component of a macromolecular machine) grossly
oversimplifies the global relationship between genotype
and phenotype.
repetitive sequences. Only ~1.5% of the human genome
codes for proteins and the remaining 98.5% includes the
other components listed above. Genotype refers more
specifically to which versions of those coding genes an
individual has.
Genome, Genotype, and Phenotype
Consequently, the central dogma is inadequate to explain
more generally how the phenotype (traits/ appearance) is
related to or derived from the genotype (genes) (Figure
1A). Every cell in a multicellular organism expresses only
a limited subset of its genes at any given time, so we must
ask whether the instructions for where and when a
particular gene should be expressed reside in the genome.
It is often stated that the genetic instructions specify the
characteristics of an organism, and that development is
simply the execution of a program encoded in the genes,
but genotype does not ‘produce’ phenotype any more than
a list of ingredients produces a pie. The signaling
molecules that turn genes on or off are regulatory proteins
encoded by other genes, but there is no genetic program
that specifies directly when or where genes are transcribed
(see Fox-Keller 2002, Ch. 4). As far as we know, the
genome does not specify when or where genes are
transcribed,
nor
how
proteins
assemble
into
macromolecular machines, nor signaling or metabolic
networks, nor any other aspect of cellular function on a
higher level of organization. During development the
genes are expressed and controlled by the metagenetic
apparatus of living cells, the biochemical machinery that
carries out molecular processes that are organized in both
specific location and direction.
The term ‘genome’ is itself a source of some confusion.
Loosely, it means ‘a complete set of genetic information’,
but when a more precise definition (‘the genes contained
on a single set of chromosomes’) is offered, its meaning
becomes less clear, especially when the sufficiency of the
genome to specify the phenotype is implied. For example,
this definition works fairly well for human females, but not
for males where genes on all 22 autosomes plus the X and
the Y chromosome are needed. Genome derives from two
greek roots –gene, meaning race or tribe, and the suffix –
ome, meaning mass or tumor. The latter connotes a
disorganized lump, collection or set, but we know that to
function properly each gene must be located near its
appropriate control sequence or cis-regulatory region.
Translocation can result in significant changes in
phenotype or cause disease, and yet ‘genome’ usually
refers to a list of the genes themselves, their positions
unspecified. This suggests a larger problem, what is lost in
the blender, the structural organization of a living cell.
In this paper, ‘genome’ includes all of the information
contained in an organism’s DNA, including the proteincoding regions, introns, control sequences, centromeres,
telomeres, genes that specify RNA products (e.g., tRNA,
rRNA, or spliceosome components), pseudogenes, and
DNA
A
RNA
Genotype
?
Protein
Phenotype
Environment
Phenotype
B
C
Genotype
P1 P2
P3 … PA
P P…
B
C
Figure 1. (A) The central dogma outlines the steps involved
in expression of a single gene, but this simple 1:1 mapping
does not explain the global relationship between genotype
and phenotype. Revised, (B) the relationship is more
complex: circular and open to signals from the environment
(see text). During development the phenotype (P1, P2 , …)
changes rapidly as growth and differentiation proceed; during
adulthood (PA, PB, …), phenotype appears static, but repair
and replacement continue (C).
‘Phenotype’, derived from the Greek phenos, meaning
appearance, refers to the organism’s discernable traits.
Here again, when one looks more closely, ambiguities
arise. Macroscopically, for example, phenotype may
appear normal (without symptoms) but a blood test reveals
abnormally low levels of a particular enzyme. Should the
results of the blood test be considered as part of the
phenotype? Within limits lower enzyme levels may not
matter, or the body may compensate, but when we describe
a phenotype in molecular and quantitative terms, we must
specify when, where, and in what amount for each protein.
During development the phenotype changes drastically
as the zygote cleaves to form a multicellular embryo,
embryonic tissues arise, and differentiation into specialized
organs takes place. The developing embryo is constructing
itself; part of this process involves shaping an environment
that is suitable for development (Figure 1B).
Control of gene expression in a given cell is triggered by
its environment –growth factors, other chemicals, and
electrical or mechanical signals. While the genotype
evokes the phenotype, genetic information does not by
itself determine what an organism becomes or does.
Instead, development of a phenotype depends on two kinds
of information: the hereditary information stored in its
DNA and the information the cell gathers (and records)
about its environment. In other words, the information
contained in a cell’s genetic code is part, but only part, of
the information that living cells process.
The insufficiency of the DNA for determining what an
organism will become –the form and appearance it takes
on during development– has been acknowledged for some
time, but recent cloning experiments have driven this point
home. Successful cloning of a domestic cat from a female
nuclear donor with calico coloring (mottled orange, white,
and black fur) produced a kitten (clone) whose coat color
does not closely resemble that of its nuclear donor “parent”
(Shin et al. 2002). This is possible because coat color
patterns are controlled during development by both genetic
and epigenetic factors. Thus, a genetic clone is not
necessarily a phenotypic copy of the donor. Furthermore,
many cloned animals survive for only a short period of
time. Shorter lifespan apparently results because long-term
survival requires that roughly half the genes must be
derived from each parent. This suggests that, in addition to
the information coded in the DNA, a genome made by
joining two haploid sets contains essential information
about ancestry, and it emphasizes the continuity from one
generation to the next. Parents contribute more than just
their DNA code.
Accordingly, an organism’s phenotype represents a
higher level than its genotype, and so in Figure 1B this is
reflected by the placement of phenotype above genotype,
not at the same level. This difference is possible because
part of the genetic code includes instructions for making
sensory devices that can detect either chemical, electrical,
or mechanical signals. Genes determine the types of
sensors a cell can make, but genes do not specify the
patterns of information a given cell will receive nor the
way it responds to any given signal. Once a cell builds and
deploys sensors, it begins to collect information (that is not
genetically encoded) about its environment. This gives it
access to periodically updated or continuous signals that
allow a primitive sense of time, and memory (cf. current
state with previous state) to develop. By recording such
signals and learning to recognize patterns that are crucial to
survival and reproduction, living organisms gain access to
a type of information that is constrained by, but not derived
from, both physical laws and genetic code. Thus, the
relationship between genotype and phenotype becomes
complex and circular, not linear, and linked to signals from
the cell’s environment (Figure 1B).
Furthermore, as multicellular bodies develop specialized
regions or clusters (e.g., an organ that performs a limited
set of tasks), there develops an obligatory dependency on
other clusters, and consequently, a heightened requirement
for communication and feedback. Accordingly, one of the
hallmarks of living systems is feedback control. Knowing
what each cluster does (its specialty) and what it needs
(what kinds of things, how much, at what rate, etc.) are
prerequisite to understanding integrated function, whether
in the developing embryo or in the adult body.
In contrast to a developing embryo, an adult’s phenotype
appears static, but this appearance is deceiving (Figure
1C). Living organisms are engaged in ongoing processes
of repair, replacement, and regeneration. Growth of hair,
claws, or nails, sloughing of dead skin, renewal of blood
cells, and protein turnover all attest to this dynamic state.
Some organisms, such as salamanders, the simple Hydra,
and most plants have a greater capacity for regeneration
than others (e.g., humans) do, but even adult human bodies
contain pockets of stem cells with substantial regenerative
capacity. Recently, the discovery that zebrafish hearts are
capable of regeneration has led biologists to rethink the
notion that terminally differentiated tissues are incapable
of significant regrowth (Poss et al., 2002).
Given the continuity of evolutionary processes, that
evolution proceeds by descent by modification of ancestral
forms, it makes sense that the processes that guide
development of evolutionarily primitive multicellular
organisms (e.g., Hydra and many protists) apply, in large
measure and with appropriate modification or embellishment, to the development of complex, highly differentiated
organisms.
At some fundamental level, then, in terms of gene
transcription, synthesis of proteins, and degradation of
damaged or obsolete molecules and cells, the processes
that build multicellular organisms are very similar to those
that maintain and repair the adult body. They involve the
same genes, operating in the same cellular context,
interacting by the same kinds of feedback controls.
References
Fox-Keller, E. 2002. Making Sense of Life. Cambridge,
MA: Harvard University Press.
Harold, F. 2001. The Way of the Cell. New York, NY:
Oxford University Press.
Leroux, M.R. and Hartl, F.U. 2000. Protein Folding:
Versatility of the Cytosolic Chaperonin TRiC/CCT.
Current Biol. 10:R260-R264.
Poss, K.D., Wilson, L.G., and Keating, M.T. 2002. Heart
Regeneration in Zebrafish. Science 298:2188-2190.
Shin, T., Kraemer, D., Pryor, J., Liu, L., Howe, L., Buck,
S., Murphy, K., Lyons, L., and Westhusin, M. 2002. A Cat
Cloned by Nuclear Transplantation. Nature 415: 859.